Birth mode
Whether an infant is born vaginally or through caesarean delivery (C-section) most drastically affects the composition of their associated bacterial communities over the first 6 months of life [
18‐
21]. For example, in a study of 98 mother–infant pairs, vaginally-born infants had less fecal
Enterobacter,
Haemophilus,
Staphylococcus,
Streptococcus, and
Veillonella species and had increased
Bacteroides,
Bifidobacterium,
Parabacteroides, and
Escherichia compared to infants born by C-section [
20]. In this same study, vertical transmission of the mother’s fecal microbiota was likely the most significant contributor to the difference between vaginally and C-section-born infants, as 72% of bacteria colonizing feces of vaginally-born infants were present in their mother’s fecal microbiome, as compared to only 41% for C-section-born infants.
Vertical transmission of fungi from mother to infant has been most extensively studied with regard to the fungal species
Candida albicans. In a study of very low birthweight infants, 24% of infants (
n = 46) were colonized at a minimum of one site (oral cavity, rectum or groin) with
C. albicans within 1 week of birth by the same
C. albicans isolate present in either their mother’s vagina, rectum, skin or mouth, as determined by culturing and DNA fingerprinting (Fig.
1) [
22]. Given this example of vertical transmission of one fungal taxon, it is likely that other members of the maternal vaginal mycobiota are also transferred to the infant. Therefore, by better understanding fungi that inhabit the birth canal of the mother, we may gain insight into which fungal taxa may be transferred to the infant.
Historically, studies have focused on the characterization of one fungal genus within the female genital tract,
Candida, due to its importance as a cause of infection (vaginitis). In a study of
Candida colonization of the vagina of asymptomatic women without a history of vaginal candidiasis, 28.8% of women were positive for
C. albicans by targeted PCR analysis, demonstrating that
Candida is a normal commensal of the female genital tract (Fig.
1). Of note, only 6.6% of these women were positive for
Candida by culture, highlighting the limitations of fungal community analysis using culture-based approaches [
5]. In another screen of asymptomatic women without a history of vaginal candidiasis, 40% (
n = 52) of the sampled individuals carried at least one isolate of
Candida (~90% of which were
C. albicans) in their genital tracts (vulva or vagina), as determined by DNA fingerprinting of cultured
Candida [
23]. To our knowledge, there has only been one published broad survey of the vaginal mycobiome using the ITS region [
24]; in this survey of healthy women (
n = 251),
Candida was the predominant genus and was present in 68% of samples, followed by
Davidiella,
Cladosporium and other less abundant fungi (Fig.
1). Of note, a large number of sequences in this study were not classified into a fungal taxonomic group, likely due to the under-representation of fungal sequences in currently available databases, as well as inconsistencies in the taxonomic classification of fungi. Nevertheless, the vaginal mycobiota likely plays an important role in the early colonization events of vaginally born infants.
Apart from
C. albicans, mother–infant transmission has not been well studied for fungi. In a recent survey of four infants and their mothers using universal 18S rDNA primers [
25], no amplicon was produced from DNA isolated from infant fecal samples at any time point, whereas the same samples were able to produce amplicons using universal 16S primers for bacteria. Fecal DNA from the mothers in this study produced 18S amplicons corresponding to
Blastocystis (a parasite of the Stramenopile group),
Saccharomyces,
Candida,
Nicotiana, and
Cicer, among other fungal genera, leading the authors to conclude that the infant fecal samples contained no fungi. In contrast, in a study of very low birth weight infants, ITS2 amplicons were produced from the feces of the majority of infant samples (7 of 11) with the predominant genus being
Saccharomyces (
S. cerevisiae) followed by
Candida (
C. albicans,
C. glabrata,
C. quercitrusa,
C. diddensiae,
C. parapsilosis, and
C. tropicalis),
Cladosporium (
C. sphaerospermum and
C. tenuissimum), and
Cryptococcus (
C. albidosimilas and
C. podzolicus) [
26]. Similarly, in a study of 11 infant fecal samples using fungal-specific PCR along with ITS2-based sequence analysis, fungi were observed in all samples, with
C. albicans being the predominant species followed by
C. parapsilosis,
C. krusei, and
Leptosphaerulina [
14].
Penicillium,
Aspergillus,
Candida,
Debaryomyces,
Malassezia,
Ascomycota,
Eurotiomycetes,
Tremellomycetes,
Nectriaceae, and
Trichosporon were also observed in an ITS1 survey of infants under 2 years of age, and the presence of many specific taxa was confirmed by culture-based methods [
27]. A longitudinal study of 14 infants also showed fungi to be present in the feces of all infants, at most time points, at a density of 10
4–10
6 rRNA genes/g of feces over the first 200 days of life [
28]. This is in contrast to the 10
9–10
10 bacterial rRNA genes/g of feces reported for the same samples [
28]. Despite one study reporting a lack of fungi in the infant GI tract, we conclude that most infants do harbor GI fungi (Fig.
1). Although fecal fungal RNA appears to be less abundant than bacterial RNA, fungal cells are considerably larger (100-fold) than bacterial cells. Thus, fungi contribute a substantial biomass to the fecal microbiota.
The extent of transfer of fungal communities from mother to infant remains unclear, but the overlap of the infant and the adult GI mycobiome does support the hypothesis of vertical mycobiota transmission. For example, in adults, fungi were detected in all fecal samples of 96 individuals by ITS sequencing, and the most abundant genera were
Saccharomyces,
Candida, and
Cladosporium [
2]. Similarly, in healthy controls (n = 12) from an ITS sequence-based survey of obese and lean individuals [
29],
Mucor,
Candida,
Penicillium,
Wallemia,
Bettsia, and
Cladoporium were the predominant genera in feces, along with more minor members, and the healthy controls (
n = 55) from a study of hepatitis-infected individuals were colonized with
Saccharomyces,
Candida,
Aspergillus,
Malassezia,
Penicillium, and an uncharacterized fungus, as determined by 18S restriction fragment length polymorphism (RFLP) analysis [
30]. The adult GI mycobiome, as characterized by ITS1 sequencing, has also been shown to contain
Candida,
Penicillium,
Aspergillus,
Malassezia,
Debaryomyces,
Mucor,
Eremothecium,
Pichia, and
Cyberlindnera (healthy controls,
n = 29) [
31]. Further,
Saccharomyces and
Penicillium species were also detected by denaturing gradient gel electrophoresis and 18S rDNA sequence analysis of fecal samples from healthy adults, along with other fungal genera [
32], and
Candida species predominated in a study of 45 adult fecal samples using ITS sequencing [
33]. In the latter study, only two fungal taxa were shared in a majority of samples, indicating an absence of core GI tract mycobiota [
33]. Additionally, the observation of instability across longitudinal samples in the study further indicated the lack of a core mycobiota, although further testing with a larger sample size should be performed to confirm this hypothesis [
33]. The fungal taxa shared across the infant and adult studies include
Candida,
Saccharomyces, and
Cladosporium, but given the limited number of studies in infants and their small sample sizes, there are likely more fungal taxa that overlap between the two groups (Fig.
1). The observation of shared taxa between adults and infants provides support for the hypothesis of vertical transmission of GI mycobiota from mother to infant.
In C-section born infants, the bacterial microbiota of the skin and GI tract are more similar to those of the mother’s skin [
21]. If this also holds true for the mycobiome, the skin and GI tract mycobiomes of C-section born infants would be expected to be dominated by
Malassezia [
34]. In a longitudinal study of infants that investigated skin colonization with
Malassezia by
Malassezia-specific PCR, two species (
M. restricta and
M. globosa, also observed on adult skin) were detected on the infant skin as early as the first day of life (89%), and abundance levels increased to that observed in adults by day 30 of life (Fig.
1) [
35]. In addition, in the same study, vertical transmission of
Malassezia from the skin of mothers to that of infants was confirmed by genotyping of the intergenic spacer region located downstream of the ITS1 and 2 regions. The targeted approach used in this study, however, prevents us from determining the proportion of
Malassezia relative to other mycobiome members, warranting further exploration of the infant skin mycobiota with broad survey approaches.
To better understand the infant skin mycobiota as a community, we can consider adult studies that employ culture-dependent and independent methods. For example, in a study of ITS1 sequences from 14 skin sites of 10 healthy adults, the skin mycobiome, with the exception of the feet, was dominated by the genus
Malassezia [
34], consistent with the findings of other studies using culture and targeted PCR approaches [
34‐
39]. Other members of the skin mycobiome, found either on a small number of people or in low abundance, included
Candida,
Rhodotorula,
Saccharomyces, and
Penicillium, along with several much less abundant fungi [
34,
37]. For studies using ITS1 sequencing, however, it should be noted that such sequences are somewhat biased toward identification of basidiomycetous fungi, such as
Malassezia, whereas ITS2 sequences favor identification of ascomycetous fungi such as most other human-associated fungi, including
Candida [
40]. Thus, it is possible that the ITS1-based approach used in the study by Findley et al. [
34] was not sensitive enough to fully detect Ascomycetes present in the samples. Nevertheless, the vertical transmission of
Malassezia from mother to infant further supports the hypothesis that birth mode impacts the mycobiota of the infant. Based on the studies mentioned above, we can hypothesize that infants born vaginally would have a higher proportion of
Candida in their mycobiome given the
Candida-predominated birth canal, and potentially have a more diverse mycobiome in comparison to those born by C-section given the exposure to the mother’s varied fecal mycobiota. Conversely, we can hypothesize that C-section infants, whose colonization source is often the mother’s skin, would be colonized by relatively higher amounts of
Malassezia (Fig.
2).
Diet
Diet composition, such as human milk or formula, strongly affects early infant bacterial microbiomes within the GI tract as well as in later infancy during the transition to solid foods [
20,
41]. For example, breast-fed infants harbor more
Bifidobacteria and
Labctobacilli in their GI tracts in comparison to formula-fed infants [
20,
28], likely due to the endogenous microbiome of human milk and human milk factors, such as oligosaccharides and immune proteins, that modulate the growth of certain bacteria. In several studies, many bacterial genera found in human milk (
Bifidobacterium,
Bacteroides,
Staphylococcus,
Streptococcus,
Pseudomonas,
Lactobacillus, and others) were also found in infant fecal samples [
42‐
44]. Additionally, prebiotics in breast milk, such as human milk oligosaccharides, promote the preferential expansion of certain taxa such as
Bifidobacterium species [
45,
46]. Although the human milk mycobiome has yet to be characterized, we know that some fungi, such as
Candida species, are found in milk from mothers with mammary candidiasis (67.4%,
n = 46) as well as in asymptomatic controls (79.1%,
n = 43) (Fig.
1) [
47]. Human milk oligosaccharides have also been demonstrated to impact fungal virulence in vitro by decreasing the ability of
C. albicans to invade intestinal epithelial cells [
48]. Thus, it is rational to hypothesize that human milk also influences the infant GI mycobiota, although this remains to be tested.
Diet also affects the oral bacterial microbiota of infants [
49], and this likely holds true for the mycobiota. In infants, oral colonization with fungi has only been studied using culture-dependent methods, with
Candida species appearing to be the predominant fungi, although this finding may be biased because these species grow most easily on certain growth media. In one study of 100 healthy infants, 12% were orally colonized with
Candida by 4 weeks of age and colonization prevalence rates did not change over the first 6 months of life [
50]. Other studies also show oral
Candida colonization rates to be low at birth, but to rise over time to adult carriage rates within the first year of life [
51,
52]. The species of
Candida currently reported as commensals of the oral cavity of infants include
C. albicans,
C. parapsilosis,
C. krusei,
C. guillermondii,
C. geocandidum, and
C. tropicalis [
51,
52]. In adults, the oral mycobiota is reportedly diverse. For example, in one study using ITS2 sequencing to characterize oral mycobiomes, healthy individuals (
n = 20) had a range of 9 to 23 fungal species within the oral cavity at a given time; the ‘core’ members of the oral community included
Candida,
Cladosporium,
Aspergillus,
Fusarium,
Glomus,
Penicillium,
Alternaria,
Cryptococcus,
Ophiosoma,
Phoma,
Schizosaccharomyces,
Zygosaccharomyces, and
Saccharomyces [
53]. Another study of adult saliva samples (
n = 3) reported
Malassezia as the dominant taxon, in addition to more minor mycobiome members such as
Epicoccum or
Irpex and others previously reported [
54]. Therefore, given the diversity of the adult oral mycobiome, we can hypothesize that the oral mycobiome of infants likely contains other taxa in addition to
Candida, and that the oral mycobiome of infants is likely dynamic, with changes especially noticeable as the infant transitions to a more adult-like diet (Fig.
2).
Gestational age at delivery
Gestational age at delivery has been shown to impact the bacterial microbiota, including initial community differences and the pace of bacterial community maturation, but differences are often resolved by 2 years of age, when the bacterial community has reached an adult-like state [
55‐
57]. For infants born at an early gestational age, the health impact of intestinal fungi is particularly significant, as the incidence of invasive, systemic candidiasis in these infants is approximately 10%, with an associated mortality rate of approximately 20% [
58]. Susceptibility to invasive infection has been correlated with relative overgrowth of fungi, especially within the GI tract [
8‐
10], as well as the presence of several predisposing clinical factors including a naïve immune system, bacterial dysbiosis due to antibiotic exposure, and use of parenteral nutrition, among other factors (Fig.
2) [
59]. In an attempt to reduce rates of invasive candidiasis in this vulnerable population, prophylactic antifungals, such as nystatin and fluconazole, are often used in neonatal intensive care units. In particular, fluconazole has been shown to reduce
Candida overgrowth at several body sites, including skin, the respiratory tract, and the GI tract, and its use has been associated with a decreased rate of invasive candidiasis in extremely low birth weight infants [
60,
61]. Recently, the enteral administration of bacterial and fungal probiotics, such as
Lactobacillus reuteri,
L. casei,
L. rhamnosus,
L. acidophilus,
Streptococcus thermophilus,
Bidiobacterium longum,
B. bifidum,
B. lactis, and
S. boulardii, have also been used to reduce invasive candidiasis [
59,
60]; however, their efficacy remains unclear and their primary site of action is limited to the GI tract. Despite the administration of antifungals, fungal colonization still occurs in some infants, as demonstrated in a study of 11 extremely low birth weight infants, all of whom received an enteral anti-fungal treatment, nystatin, as well as antibacterial antibiotics [
26]. Of these 11 infants, 7 produced ITS amplicons from their stool, with
Saccharomycetales being the most abundant order, as well as
Candida and
Cryptococcus species. Thus, despite efforts to prevent fungal colonization, fungi maintain the ability to persist in the infant GI tract. In infants born at early gestational ages, beneficial fungi, such as
S. boulardii, may help to regulate the growth of opportunistic fungal colonizers such as
Candida. Additionally, given the association of bacterial dysbiosis with systemic candidiasis in infants born at an early gestational age, a robust bacterial community may play an important role in mycobiota regulation, as discussed below.